Image Sharing Suggestions via Probabilistic Facial Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image management systems fail to suggest sharing of images containing people important to users, as they do not accurately recognize relative importance of individuals in images and do not learn thresholds based on user-specific image account composition.
Innovation Solution
A computer-implemented method using a probabilistic model to determine the share probability score of images based on facial recognition, cluster rank values, and user account composition, providing suggestions for images with people of relative importance to the user.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional image management systems provide sharing suggestions for all images, then the quantity of sharing suggestions increases, but the relevance and user acceptance rate decreases due to inability to accurately recognize relative importance of individuals
Solution Approach 1:
The system changes the parameter of importance recognition by using facial recognition technology to identify and rank individuals in images. It calculates cluster rank values based on multiple parameters including face quality, time period counts, image counts, recency, and name labels associated with each person cluster, thereby accurately determining the relative importance of individuals depicted in images.
Solution Approach 2:
The patent replaces manual assessment of image sharing importance with an automated computational system. The probabilistic model automatically calculates share probability scores by processing facial recognition data, cluster rankings, and account composition information, substituting human judgment with machine learning-based automated decision-making.
2Ease of operation
If the system provides sharing suggestions without normalizing for account size, then processing simplicity is maintained, but the relevance of suggestions deteriorates across users with different account compositions
Solution Approach 1:
The system applies local quality by normalizing the share probability score specifically for each user's account size. It calculates a normalization factor based on the ratio of the user's account size to the average account size, and multiplies this factor with the raw share probability score. This tailored approach ensures that suggestions are appropriately adjusted for each user's specific account composition rather than applying a uniform threshold to all users.
3Measurement precision
If the system uses multiple parameters for cluster ranking, then the accuracy of importance determination improves, but the device complexity increases
Solution Approach 1:
The system segments the complexity by organizing multiple ranking parameters into distinct, modular components. Each parameter (face quality, time period counts, image counts, recency, name labels) is processed separately to generate cluster rank values, which are then integrated into the overall importance determination. This modular segmentation makes the complex multi-parameter system more manageable and computationally efficient.
Data Source
AI summary
Some implementations can include a computer-implemented method and/or system for automatic suggestions to share images containing people of importance to a user. The method can include determining, based on pixels of an image associated with a user account, one or more clusters associated with the image. The method can also include determining a share probability score for the image based on a probabilistic model and determining that the share probability score meets a threshold. The method can further include, in response to determining that the share probability score meets the threshold, providing a suggestion to a user associated with the user account to share the image.